History of Soy Oil Hydrogenation and of Research on the Safety of Hydrogenated Vegetable Oils - Page 3

A Special Report on The History of Soy Oil, Soybean Meal, & Modern Soy Protein Products

A Chapter from the Unpublished Manuscript, History of Soybeans and Soyfoods: 1100 B.C. to the 1980s

by William Shurtleff and Akiko Aoyagi

Copyright 2007 Soyinfo Center, Lafayette, California

Page 1 | Page 2 | Page3

In 1976 Ahrens (at the Rockefeller University), a highly respected and experienced researcher in this field, pointed out in an excellent article, that the subject of serum cholesterol levels is a serious one, since it is an undisputed fact that people with higher serum cholesterol levels have more and earlier heart attacks than those with lower cholesterol levels, as first shown by the famous Framingham study. Yet as he also pointed out, there has never been rigorous proof of the Lipid Hypothesis, namely that lowering the level of blood cholesterol in an individual or population will lead to a reduction in the risk of suffering a new event of coronary heart disease. Because of this lack of convincing proof on humans (much evidence from animal experiments supports the Lipid Hypothesis) and because of the difficulties in getting people to change their diet in a way that significantly lowers serum cholesterol, Ahrens noted that "The medical body politic is not convinced today that lipid-lowering in the general public is either feasible or effective in reducing the incidence of coronary heart disease; nor am I, though I suspect we all have high hopes." Ahrens was then actively involved in testing the Lipid Hypothesis and in teaching patients "how to live comfortably and happily on diets low in saturated fat and cholesterol, high in polyunsaturated fat--and how to stay on these diets for years." In 1979 Ahrens gave an update on his views, again provocative.

In 1976, in an extensive review of the literature, the Federation of American Societies for Experimental Biology concluded that:

Although the results of studies on the effect of trans-isomers on serum lipid levels in humans are not definitive, the weight of evidence indicates that trans-monoenoic acids, the principal geometric isomers presented in hydrogenated soybean oil, are not hypercholesterolemic. Similarly, the results of animal experimentation indicate that trans-acids of hydrogenated soybean oil are not atherogenic at normal dietary levels.

In 1979 in a PhD dissertation, C. Moore concluded: "Results of this investigation suggest the large amounts of trans -fatty acids in the American diet do not increase serum cholesterol." Likewise, in an authoritative review, Emken (1980a) concluded: " . . . human studies indicate partially hydrogenated soybean oil produces little or no increase in the rate of atherosclerosis development. In fact, the polyunsaturates remaining in partially hydrogenated soybean oil may be exerting a beneficial effect."

In 1981 Applewhite did the most detailed analysis to date comparing the key variables in 13 studies on trans fats and human serum cholesterol levels, including all those cited above. He argued quite convincingly that wherever trans had apparently been shown to raise human serum cholesterol levels there was a fault in the experimental design, often a shortage of essential fatty acids.

Two animal studies done by Decker and Mertz in 1966 and 1967 have been widely cited as indicative of biochemical changes and possible health hazards attributable to trans fatty acids. The first study showed that animals fed a low percentage of trans fatty acids accumulated these in the red blood cells and mitochondria (energy producing structures within the cell). The second reported that this accumulation contributed to an increased rate of mitochondria swelling and erythrocyte hemolysis (red blood cell disintegration or dissolution), the hemolysis rate being five times that of the control cis group. In criticizing these studies Applewhite (1981) has noted that the olive oil used in both these studies was not hydrogenated but rather elaidinized with sulfur dioxide, which probably reduced the essential fatty acid content to marginal amounts (as the authors acknowledged) and introduced a relatively large amount of mixed linoleic acid isomers into the diet. Applewhite noted that research done in 1964 showed that EFA deficiency leads to increased hemolysis of red blood cells. Subsequent, more carefully designed studies by Emken and others have shown no evidence of permanent or preferential accumulation of trans fatty acids in human tissues. The studies by Decker and Mertz and others in the late 1960s led to a growing interest in the metabolic fate of fatty acid isomers formed during hydrogenation and in the effect of trans on the biophysical properties of cell membranes.

Several other lines of research in the late 1960s were influential. In 1967 Scholfield et al. at the USDA Northern Regional Research Center, using liquid chromatography, demonstrated vividly the large number of geometric and positional isomers created during partial hydrogenation, and developed a simple graphical technique for indicating the distribution of these in any oil sample. Holman (1968^??) and three other researchers in the early 1970s showed that trans polyunsaturated fatty acids are not converted into biologically active prostaglandins as typical cis polyunsaturates are.

Starting in the 1960s and increasing during the 1970s, a dimension of politics and vested economic and personal interests began to play an important role in the research on the safety of hydrogenated products. This was also a period during which there was a switch from university funding of university research to outside and especially corporate funding. After US consumption of margarine passed that of butter in 1957 and as vegetable shortening raced ahead of lard, while per capita consumption of both butter and lard fell steadily, the US dairy and meat industries began to sponsor their own research into the safety of their competitors' products. If hydrogenated products were, in fact, unsafe, the dairy and meat industries wanted to be sure that consumers got the message. (The meat and dairy interests could not, however, suggest that trans fatty acids might be harmful, since their products also contained trans !). Likewise the manufacturers of hydrogenated oils, margarines, and shortenings, each with high investments to protect, hoped that the research they sponsored would show that their products were safe and nutritious. Although one would like to believe that science is above politics, it is still worth noting that, for example, the influential research of Deuel and Alfin-Slater was funded by Best Foods Inc., makers of many best-selling hydrogenated products. Only a few institutions (perhaps the USDA NRRC is the best example) and individuals were not at least subtly aligned on one side or another of the increasingly heated issue.

An additional dimension was that researchers looking for grants and funding or trying to get papers published had an interest in finding problems with hydrogenated products. Since the body is quite sensitive to increased levels of fats, it was easy for researchers who wished to obtain almost any results they wished by simply designing their experiments in a certain way. They could "load the dice" by creating an experimental diet vastly different from the normal diet, overloading the system with excess elaidic acid or highly saturated fats (which can force the body to use these substances in ways it would not ordinarily use them) or feeding very low levels of essential fatty acids. Likewise, pro-hydrogenation factions could use low enough levels of problematic ingredients that even if there were a real problem it would not show up in the experiment. Moreover, during the 1970s and 1980s, as new scientific studies or governmental or committee reports were published concluding that hydrogenated fats were either probably safe or possibly unsafe, the "opposing camp" would almost routinely give their own critique of the evidence, attempting to show that it was incorrect. (One of the few groups above these vested-interest conflicts were USDA researchers.) Understandably the public became increasingly confused, and some of the media played into the hands of commercial interests they would not ordinarily side with, largely from lack of awareness of the political forces behind apparently scientific debates. Thus it became increasingly important for anyone attempting an objective understanding of the increasingly complex subject of the safety of hydrogenated oils, to at least understand who was funding the research and how the experiment was designed.

Two groups of researchers whose work was most widely criticized by the hydrogenated oil interests was the group headed by Kummerow at the Burnsides Research Laboratory (affiliated with the University of Illinois) and the Keeney-Enig group at the University of Maryland. Both were funded primarily by the National Dairy Council and the National Livestock and Meat Board, and both obtained wide publicity for their anti-hydrogenation publications. The hydrogenated oil industry spent considerable effort in finding fault with their experimental designs and refuting their findings. Dr. Kummerow spent many years in trying to prove that hydrogenated fats were unhealthy. His early work in the late 1950s and 1960s, cited previously, was not particularly controversial. In 1970 Sgoutas and Kummerow reported that trans fatty acids differed from their cis counterparts in the manner in which they were incorporated into triglycerides and phospholipids and that they "seem to alter the permeability characteristics of cell membranes, which may result in changes in their biological function." Then in 1974 Kummerow and his group published a study on swine feeding, which showed that feeding young swine 19% dietary partially hydrogenated soy oil (iodine value = 65) with 50% trans content caused the greatest elevation of total lipid and cholesterol levels and the largest area of aortic atherosclerosis of the various diets fed. A press conference was held and the findings received extensive and dramatic media publicity. Headlines questioned the safety of margarine. When the findings were presented at two professional conferences in 1974, they were questioned and the suggestions that the results should be attributed to the effect of a high fat-to-protein ratio and borderline deficiency of essential fatty acids were acknowledged by Kummerow in print in 1975. In June 1974 the National Association of Margarine Manufacturers issued a detailed critique and refutation of Kummerow's findings on swine atherosclerosis. Subsequent feeding studies of young swine at the Universities of Wisconsin (Elson et al. 1976^??) and Illinois (Jackson et al. 1977^??) using lower fat-to-protein ratios and increased EFA levels with dietary fats containing up to 45% trans isomers failed to substantiate Kummerow's results. In 1977 Hsu and Kummerow reported that elaidic acid (9t-18:1, the main trans isomer from hydrogenation) is oxidized by the heart muscle mitochondria less effectively than vaccenic acid (11 t- 18:1, the main trans acid in animal fats) or than cis -isomers of oleic acid (9 c- 18:1), and may therefore affect the ability of the heart to respond to stress. The latter statement was speculative. In 1977 Perkins and others (with Kummerow) investigated the fatty acid composition of 12 commercial margarines. The content of 18:1 trans ranged from 6.9-30.8% by weight and averaged 18.0%, while the content of 18:2 trans-trans and cis-trans ranged from 0.7 to 8.6% and averaged 3.2%. In two samples of human milk 18:1 trans was 2.1 and 4.0% by weight and 18:2 trans was not detected. In human blood 18:1 trans comprised 1.9% of the fatty acids in the serum and 2.4% of those in the red blood cells. The figures for 18:2 trans were 0.8% and 0.7% respectively. In his numerous articles and speeches, Kummerow continued to berate hydrogenated oils and criticize all recommendations that Americans should reduce their consumption of meat and of animal fats (Kummerow 1979). A summary of Kummerow's views was given in 1979 in a chapter in Geometrical and Positional Fatty Acid Isomers , edited by Emken and Dutton. There he set forth in a new form his thesis that trans acids cause coronary heart disease (CHD), noting that southern Europeans (e.g. from Spain, Italy, Yugoslavia, Greece) who consume relatively little hydrogenated fats have low rates of CHD, as do Bulgarians, who consume lots of saturated fats and cholesterol, but little hydrogenated fats.

As mentioned earlier, the other group, supported by dairy industry funds, that studied the safety of hydrogenated oils, was the Keeney-Enig group at Maryland. In 1978 and 1979 Enig, Munn, and Keeney^??, in various articles, did retrospective analyses on data concerning fat consumption and cancer as one way of "providing clues to the etiology of these diseases." They reported that the percentage of total fats in the American diet that came from animal fats declined steadily from 83% in 1909 to 62% in 1972. During that period total per capita daily fat intake increased from 125-156 gm. Most of the increase came from vegetable fats and saturated fats accounted for only 19% of the increase. However the major change in fat consumption during the 20th century was the increase in partially hydrogenated fat consumption in the form of margarines, salad and cooking oils and shortenings. They estimated that trans fatty acid content of the diet during that period increased from an estimated 4.4-12.1 grams per day, and in the mid-1970s accounted for an estimated 8% of total fat consumption and 20.5% of vegetable fat consumption. They then reported positive correlations between the dietary intake of hydrogenated vegetable fats that contain trans fatty acids and total cancer mortality as well as breast and colon mortalities, and negative correlations with the same death causes and total animal fats. In a lively exchange of views, this report was criticized by representatives of the margarine and shortening industries and by the editor-in-chief of the Journal of the National Cancer Institute for factual errors, for using correlations to imply cause and effect and thus taking a simplistic approach to a very complex problem, and for using only selected data in comparing cancer rates in certain countries that led to desired conclusions (Applewhite et al. 1979 ref??, 1981). Acknowledging several factual errors, the authors rebutted that they were not trying to prove cause and effect but simply "hoped to identify an alternative body of evidence" as to what is causing cancer, on which more research should be done.

As the stakes for the dairy and hydrogenated oils industries grew bigger and the issues grew more political, they also grew more complex and subtle, so that eventually only a relatively small body of experts could even understand the language in which the debate was taking place. The use of new equipment, such as gas liquid chromatography in place of the traditional infrared spectrophotometry, made it possible to detect smaller concentrations of trans fatty acids and other substances in smaller samples. As scientists came to better understand the many isomers created by hydrogenation of an oil but not found in the natural oil, the intricacies and subtleties of the human metabolic process, and the complex interaction of these two, many new questions and unknowns appeared.

In order to deal with the expanding literature on the subject, to bypass politics, and to translate the basic findings into conclusions applicable to the realities of daily diets, objective reviews and evaluations of the research findings became necessary. The first of these was a 94-page review of the scientific literature from 1920-1973 related to the safety of hydrogenated soy oil as a food ingredient. Containing 430 references, the report was done for the US Food and Drug Administration by Tracor-Jitco, Inc. Despite its scope, the report made no conclusions or recommendations, and did not cover virtually any of the live issues related to the concerns with trans fatty acids. It seemed to imply that hydrogenated soy oil was a safe food ingredient. In 1976 the Select Committee on GRAS (Generally Recognized as Safe) Substances of the Federation of American Societies for Experimental Biology prepared a 40-page report entitled Evaluation of the Health Aspects of Hydrogenated Soybean Oil as a Food Ingredient (FASEB 1976). The committee of eleven eminent scientists reviewed the pertinent literature (90 references) and concluded: "There is no evidence in the available information on hydrogenated soybean oil that demonstrates, or suggests reasonable grounds to suspect, a hazard to the public when it is used as a direct or indirect food ingredient at levels that are now current or that might reasonably be expected in the future."

In 1979 the first full book on trans fatty acids was published; clearly the subject was attracting growing interest and attention. Edited by Emken and Dutton, both highly respected researchers from the USDA NRRC, Geometrical and Positional Fatty Acid Isomers (344 p.) contained chapters by eleven of the top researchers in the field, representing a variety of opinions It clearly reflected the growing complexity of the subject and growing number of unanswered questions. Various authors showed that trans fatty acids may be metabolized quite differently from their cis counterparts and that both geometric and positional isomers are recognized by cellular systems and treated differently than cis isomers. Emken, for example, concluded his excellent and objective chapter on "Utilization and Effects of Isometric Fatty Acids in Humans" with the statement: "The fact still remains that we know far less than we should about the metabolism of this major dietary food." Dutton in a personal interview agreed: " Trans acids are a big unknown in metabolic processes. The subject must be researched more." Emken later (1980a) pointed out that little is known about the effect of trans acids on the more than 28 disorders (which he listed) that involve lipid metabolism. He concluded:

There are still many unanswered questions concerning the nutritional effects of hydrogenated and heated oils. Relatively little information is available concerning the biological threshold at which partially hydrogenated or heated fats can be tolerated without affecting our nutritional well-being. A major difficulty is to separate effects due to fatty acid isomers and products in heated fats from effects due to the various normal physiological influences of the high fat content of the American diet. A second difficulty is the identification of genetic traits which can greatly influence human nutritional response to partially hydrogenated or heated fats.

In 1980 the Ad Hoc Committee on the Composition of Special Margarines (Brisson 1980??), advising the Health Protection Branch of the Canadian government issued a special report, initially in summary form. The report made a number of specific recommendations and expressed various concerns. Recommendations included (1) permitting labeling of the cis-cis linoleic acid content of oils, shortening, and margarines when that content was relatively high, as an aid to those wishing to lower their serum cholesterol; (2) encouraging the industry to increase the linoleate content while decreasing the trans fatty acid content of Canadian foods; (3) regulating all margarines and shortenings to require less than 1% of 18:2 trans-trans fatty acids; (4) limiting the content of 18:1 trans fatty acids in special margarines to about 15% of fatty acid content; and (5) limiting the content of saturated plus trans fatty acids in special margarines to no more than 40%. This Committee was the first high-level, scientific, and presumably impartial body to bring the various concerns over the safety of trans fatty acids to public attention. "The Committee encountered many areas of uncertainty with respect to the knowledge of metabolism of trans acids and of polyunsaturated fatty acids and their derivatives. We strongly recommend further research in this field, supported by government, industry, and universities." They also showed concern over the rising levels of trans fatty acids in human diets and over the unique way in which trans isomers were metabolized by humans and experimental animals, a way similar to that of saturated fats. Yet concerning trans fatty acids they did note that " Trans-trans -18:2 excepted and with adequate linoleate intake (i.e. essential fatty acids), there is no strong evidence that they are any more harmful than any other fatty acid class."

The Canadian meat and dairy industries were put in a difficult position by the report. On the one hand they liked the disclosure of concerns over the high levels of trans in hydrogenated fats, but they disliked the recommendations that made the essential fatty acid linoleic acid look too health-promoting lest people decrease their consumption of foods such as butter and red meats, which contained little linoleic acid. In June 1981, at the request of the Dairy Bureau of Canada, a "Committee of Medical and Scientific Experts" (chaired by Brisson and Keeney, with the assistance of Enig and Kummerow, among others) published what came to be known as the "Brisson Report," a critique of the Canadian Ad Hoc Committee's Report. It argued, among other things, that there was no need for special margarines labeled on the basis of their content of polyunsaturated or essential fatty acids (EFA), that no evidence existed of an EFA deficiency in humans, that no investigations had shown that linoleic acid, the EFA, is in any way a preventive factor in coronary heart disease, that high levels of polyunsaturated fatty acids have not yet been proven beneficial (and unnecessarily high levels may actually be harmful), that the Lipid Hypothesis has not been proven, and that the main trans fatty acid in animal products is different from that found in hydrogenated fats, and is less abundant. The authors then gave a good summary of ten primary and secondary intervention studies done on large groups of people between 1965 and 1978 in which major dietary changes had lowered serum cholesterol by an average of only 10% and had had no effect on mortality. Brisson then wrote a book (ref??) incorporating his Committee's critique, which was criticized by the press. In October 1981 Walter Meyer of the Institute of Shortening and Edible Oils in Washington, D.C., wrote a critique of the Brisson critique, which also criticized the recommendations and concerns of the original Ad Hoc Committee on Special Margarines. Clearly the subject of the safety of hydrogenated oils was loaded with disagreements among experts, strong feelings, and politics.

In September 1980 the American Soybean Association sponsored a symposium on the Hydrogenation of Oils in Rimini, Italy. The proceedings were edited by Leysen (1980). There was considerable discussion of health-related issues, problems of too little or too much linoleic acid in the diet, trans acid reducing the permeability of and stiffening cell membranes, and the fact that many questions were raised that could not be answered.

In March 1981 Applewhite gave one of the most detailed and carefully reasoned defenses yet of the position that there is nothing wrong with trans fatty acids. A high-level employee of Kraft Inc. (which sells both dairy products and hydrogenated oil products), former president of the American Oil Chemists Society, and chairman of the National Association of Margarine Manufacturers' Technical Committee, Applewhite knew his subject well. Although he sidestepped many of the difficult metabolic questions raised in the book edited by Emken and Dutton in 1979 and oversimplified to the point of almost misrepresenting Ahrens views on the Lipid Hypothesis and the question of the use of the term "unnatural" as applied to fatty acids by many researchers (both discussed above), Applewhite did deal very effectively with a number of the main issues concerning the safety of trans , and especially the issue of elevation of human serum cholesterol levels (as discussed earlier). His main conclusion was that there is no firm evidence showing any significant health problems with trans or with hydrogenated oils. What Applewhite failed to point out, however (as Emken noted in an 1982 personal communication), is that there are still a number of important unanswered questions related to trans , and until those answers are known, it cannot be said for sure that hydrogenated fats are safe.

Several other research reports done during the 1970s bear mention. Dutton (1974) reported that unhydrogenated liquid oils (including soy oil) that had been winterized and/or deodorized contained 2.7-5.4% trans fatty acids. These were apparently produced during refining. Heckers et al. (1979^??) and Ohlrogge et al. (1981^??) demonstrated that trans isomers do not accumulate in human tissues but instead are metabolized at a rate consistent with that of other fatty acids. There was growing interest in the effects of trans isomers on prostaglandin synthesis; prostaglandins are potent bioregulators that can control or alter a multitude of physiological functions.

To summarize: By the late 1970s, the debate on the safety of hydrogenated oils for food use had boiled down largely to a debate on the safety of trans fatty acids. It seems to us that the extensive body of scientific research on this and related subjects, based on more than 50 years of investigation, can be viewed from either of two basic perspectives. The first states essentially that, given the large body of research done to date, hydrogenated oils should be considered safe until clearly proven unsafe. When we interviewed top (and we felt nonpartisan) researchers in this field to ascertain their personal views, the typical reply was: "There are no reports in the literature which show conclusively that hydrogenated oil leads to health problems in human subjects, a number of expert committees have deemed them safe, and 75 generations of rats have thrived in good health on a diet containing hydrogenated fats as virtually its only source of fats. The subject of trans is still a very open question and there are many unknown, so we must keep a careful eye on new developments. Yet I feel no hesitation about consuming hydrogenated fats in my diet in moderation."

The second perspective, held by some people interested in natural foods, is a more cautious one. It states that since the majority of trans fatty acids produced by hydrogenation (i.e. elaidic acid) as well as many of the less prominent isomers are found in only trace amounts in natural foods??, since each of these isomers is known to be metabolized by the body slightly differently, since there are still many unknowns concerning the safety of trans fatty acids (such as effects of prostaglandin metabolism or cell membrane function), since most of the studies have been done on animals and their significance to humans is not clear, and since the typical layman cannot possibly keep up with or evaluate the literature in this increasingly complex, technical, and political field, it is probably safest to try to minimize one's consumption of man-made trans fatty acids, while eating a diet that is low in total fats, saturated fats, and cholesterol, but relatively rich in essential and polyunsaturated fatty acids, but without too much of the latter, which in excess may be carcinogenic.



Page 1 | Page 2 | Page3